So its done! I finally released a fairly easy to install (from scratch) Hugo based Static Site generator with full ActivityPub support.
It has step-by-step instructions on how to set it up for your own blog or static website.
One of the coolest features for me, other than having your static site blog posts show up as posts in the Fediverse is the support for interacting with those posts. Any replies you leave, likes, or boosts will show up in the "comments" section of the website on the page associated with the post. How cool is that!
@ephemeromorph Wonderful, let me know if you need any help going through it. I think i am the only one to deploy it so far, though I've done it 3 times.
1❓Do you know anyone willing to set up a website to do this (for money or not setup basically a Hugo static site + Fedi interaction while being privacy-minded / reduce using google type stuff / building it with less wanting it for mass-scraping so easily).
2 ❓Did you manage to replicate / get your own static site going with Fedi stuff added?
And here I am trying to figure out the easiest way to flush my server's block list without destroying the server's database...
I could do it via SQL query, but I'm not certain that's ALL that the server software needs done...
Been messed up before like that... Cleared the addresses, but there were other references and that software never worked the same. Ended up wiping and starting over.
It'd be nice if we had clear documentation for using the cli and scripting it... I just haven't found it yet.
@paul I found it to be a arelatively simple static site engine overall. But it does take some time to learn the language. If there is something in particular that confuses you just let me know.
@freemo I am doing this as a new project without fedipage.
First I want to get a bit of a feel for it. If it's easy enough, I may examine if there is a way to import my various existing (static site) sites into Hugo, through some converter or so. Those have hundreds of articles and pages, so I'm not typing them in again 😉
Also I'm not going to put a lot of time into it every day. My workdays are long enough. 😉
I moved over from middleman static site generator. The pages moved over without a hitch for me. Any markdown based source should be relatively easy to convert.
Keep in mind Hugo wont give you fediverse/activityPub access on its own. Its just another (nice) static site generator. The ability to integrate with the fediverse is entirely a fedipage thing.
@freemo My sites are in a SQLite database, created by Publii, which creates clean HTML when I output the site. Nothing markdown about it.
This does not sound like an easy transfer, then. Thank you for the information.
I'm not focussing on the Fedipage part for anything at the moment. If I can't make Hugo work for me 'as is' (including importing my existing sites), then that route is not happening.
It might be interesting for a new site I could set up. But that's not in the books yet.
@paul Ohhh its a wysiwig... yea sadly unless you can find a way to convert it you may be in for some trouble. Personally this is why i dont use blogging tools that dont use open standard formats like markdown to begin with.
Hugo does support html markup but you'd likely have to heavily edit that format since it would have to be an html fragment without all the header and styling stuff.
@freemo Done messing for now. I have 2 posts in the posts directory and the localhost:1313 only shows an empty page, with source
<pre>
</pre>
So that's very Zen but not what I am looking for. 🤣
@paul haha its fairly automatic if you have the right directory structure and settings. Can be hard to do from scratch though if you dont know the right settings.
@mur2501 Aparapi is one of my most challenging projects to contribute to. but also very rewarding.
It is also relatively active with a few devs and with my main company getting funding it will likely funnel some resources into Aparapi. So its a great project to get involved in especially right now.
@freemo
for the documentation part where you try to explain JVM Bytecode I think preceding it with Dennis Yurichev's Reverse Engineering Book's Chapter 4 on JVM would be a good idea to make the doc more consumable for people looking at JVM Bytecode for the first time in life.
@freemo
Change in plans it looks like Yurichev has paywalled his open source book. I am confused about the legalities of directly copying content (with attribution ofcourse) :blobcatsadlife:
@freemo Well my pirated pdf copy says CC licensed on the front (which technical means it is not pirated)
It is behind a paywall but the license is still CC
@mur2501 Ok then he cant paywall it.. as long as you keep the CC license attached to his work, AND you use the same license on any derrivative works then you are free to share it and legally bypass his paywall.. he is allowed to charge, but you are allowed to give away for free.
@mur2501 I wouldnt link it no... I would copy it to our server and link to it on our own domain. This way they cant paywall it or cut off access (which they cant legally do once they open source it)
@freemo
I have looked into the problem it seems the main issue is using http for googleapis rather than https, browsers seem to block all http link by default. I will change them to https links.
@freemo I have committed but haven't pushed as there are some other errors and stuffs to due the code being 7 years old. Alot of the links are dead. I have removed the Google Plus button as it exists no more.
Twitter works but the handle is it active still?
and yes in the demonstration page many websites seem to exists no more.
PS: Also I think it would better if we use CDN links rather than actual JS files
@freemo just noted a bug here regarding the visibility of your post on the *keys: the link to fedipage does not show up for some reason > screenshot from blahaj.zone > same result on firefish, misskey etc
@testing Not a bug.. i changed some things about how the site renders its pages so it broke old boosts knowingly.. but it continues to work properly.. thats just a "ghost"
@testing Thank you. Let me know if i can help at all..
Keep in mind being a bit new if you follow it you may get occasional repeat notifications... I try to keep it minimal though. This only happens when i need to change the way AP renders and it needs to resend all the pages out... should be fairly rare but might happen once or twice as i improve the system.
So your welcome to follow but please do so at your own risk :)
@freemo I love hugo. Probably should try new version, not the 5y old one currently using.
On the other side it's a beauty of static hugo binary - even 10y old could still generate web.
I dont use vercel normally either. I used vercel for the first time for this project and liked it.
To answer your question it depends if you want full features or not.
Witout vercel you wont have the serverless functions. They handle things like accepting new followers. So without vercel (or something to take its place) then people can see your account via the fediverse but following it wouldnt really work (you wouldnt see new posts in your feed and your follow request will never get accepted)... but otherwise it will "work" with limited functionality.
That said the serverless functions can be moved to any platform.
You could in theory get rid of the serverless functions and make it all truely static. If you did that (which would take a bit of coding) then theonly time followers would get confirmed would be when you push a new version of the site.. but otherwise you could, in theory, get the full functionality out of it. with some creativity.
Add comment